<div class="gmail_quote"><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">I thought Pavel's suggestion was just fine and, indeed, the result is more than one item (transcript below).</blockquote>
<div><br>Something must be lost in my translation from my real data to this simplified example, because it definitely doesn't work with my real data. I only get 1 result.<br> <br></div><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
Find another processor?<br>
</blockquote><div><br>I'm not sure which processor exist-db 1.2.5 is using out of the box and if it can be swapped.<br> <br></div><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
BTW, your data set happens to have steadily increasing version values for each <id>. If it is true that the last of each <id> has the highest version, this might execute even faster because it has no function call:<br>
<br>
/items/item[not(id=following-sibling::item[1]/id)]<br>
</blockquote></div><br>Thanks, but the real data isn't guaranteed to be in this order. It was just a simplified example.<br><br>Thanks for the responses though!<br clear="all"><br>-- <br>A. Steven Anderson<br>Independent Consultant<br>
<a href="mailto:steve@asanderson.com">steve@asanderson.com</a><br>